1. Field of the Invention
The present invention relates to a gas sensor for measuring oxides such as NO, NO.sub.2, SO.sub.2, CO.sub.2, and H.sub.2 O contained in, for example, atmospheric air and exhaust gas discharged from vehicles or automobiles, and inflammable gases such as CO and CnHm. In particular, the present invention relates to a gas sensor for measuring NO and NO.sub.2.
2. Description of the Related Art
Exhaust gas, which is discharged, for example, from vehicles or automobiles such as gasoline-fueled automobiles and diesel powered automobiles, contains nitrogen oxides (NOx) such as nitrogen monoxide (NO) and nitrogen dioxide (NO.sub.2), as well as carbon monoxide (CO), carbon dioxide (CO.sub.2), water (H.sub.2 O), hydrocarbon (HC), hydrogen (H.sub.2), oxygen (O.sub.2) and so on. In such exhaust gas, about 80% of the entire NOx is occupied by NO, and about 95% of the entire NOx is occupied by NO and NO.sub.2.
The three way catalyst, which is used to clean HC, CO, and NOx contained in the exhaust gas, exhibits its maximum cleaning efficiency in the vicinity of the theoretical air fuel ratio (A/F=14.6). If A/F is controlled to be not less than 16, the amount of produced NOx is decreased. However, the cleaning efficiency of the catalyst is lowered, and consequently the amount of discharged NOx is apt to increase.
Recently, in order to effectively utilize fossil fuel and avoid global warming, the market demand increases, for example, in that the discharge amount of CO.sub.2 should be suppressed. In order to respond to such a demand, it becomes more necessary to improve the fuel efficiency. In response to such a demand, for example, the lean burn engine and the catalyst for cleaning NOx are being researched. Especially, the need for a NOx sensor increases.
A conventional NOx analyzer has been hitherto known as an instrument for detecting NOx. The conventional NOx analyzer is operated to measure a characteristic inherent in NOx, based on the use of chemical luminous analysis. However, the conventional NOx analyzer is inconvenient in that the instrument itself is extremely large and expensive. The conventional NOx analyzer requires frequent maintenance because optical parts are used to detect NOx. Further, when the conventional NOx analyzer is used, any sampling operation should be performed for measurement of NOx, where it is impossible to directly insert a detecting element itself into a fluid. Therefore, the conventional NOx analyzer is not suitable for analyzing transient phenomena such as those occur in the exhaust gas discharged from an automobile, in which the condition frequently varies.
In order to dissolve the inconveniences as described above, there has been already suggested a sensor for measuring a desired gas component in exhaust gas by using a substrate comprising an oxygen ion-conductive solid electrolyte.
FIG. 14 shows a system of a gas analyzer disclosed in International Publication WO 95/30146. This apparatus comprises a first chamber 4 into which a measurement gas G containing NO is introduced through a narrow hole 2, and a second chamber 8 into which the measurement gas G is introduced from the first chamber 4 through a narrow hole 6. Wall surfaces for constructing the first and second chambers 4, 8 are composed of partition walls 10a, 10b made of zirconia (ZrO.sub.2) capable of transmitting oxygen ion. A pair of measuring electrodes 12a, 12b and a pair of measuring electrodes 14a, 14b for measuring the partial pressure of oxygen in the respective chambers are arranged on portions of one ZrO.sub.2 partition wall 10a corresponding to the first and second chambers 4, 8 respectively. A set of pumping electrodes 16a, 16b and a set of pumping electrodes 18a, 18b for pumping out O.sub.2 in the respective chambers to the outside of the chambers are arranged on the other ZrO.sub.2 partition wall 10b.
The gas analyzer thus constructed functions as follows. Namely, the partial pressure of oxygen contained in the measurement gas G introduced into the first chamber 4 through the narrow hole 2 is detected by a voltmeter 20 as an electric potential difference generated between the measuring electrodes 12a, 12b. A voltage of 100 to 200 mV is applied between the pumping electrodes 16a, 16b by the aid of a power source 22 so that the electric potential difference is adjusted to have a predetermined value. Accordingly, O.sub.2 in the first chamber 4 is pumped out to the outside of the apparatus. The amount of pumped out oxygen can be measured by using an ammeter 24.
The measurement gas G, from which almost all O.sub.2 has been removed, is introduced into the second chamber 8 through the narrow hole 6. In the second chamber 8, an electric potential difference generated between the measuring electrodes 14a, 14b is detected by a voltmeter 26. Thus the partial pressure of oxygen in the second chamber 8 is measured. On the other hand, NO contained in the measurement gas G introduced into the second chamber 8 is decomposed as follows by the aid of a voltage applied between the pumping electrodes 18a, 18b by means of a power source 28: EQU NO.fwdarw.(1/2)N.sub.2 +(1/2)O.sub.2
O.sub.2 produced by the decomposition is pumped out to the outside of the second chamber 8 by the aid of the pumping electrodes 18a, 18b. A value of an electric current generated during this process is detected by an ammeter 30. Thus the concentration of NO contained in the measurement gas G is measured.
However, in the case of the gas analyzer constructed as described above, if the concentration of oxygen contained in the measurement gas G is high, O.sub.2 in the first chamber 4 cannot be sufficiently pumped out to the outside of the chamber 4 by the aid of the pumping electrodes 16a, 16b. As a result, unprocessed O.sub.2 enters the second chamber 8 together with NO. Therefore, an error due to O.sub.2 occurs in the current value obtained by decomposition of NO.
Namely, as shown in FIG. 15, the pumping electrode 16b in the first chamber 4 is arranged only on one ZrO.sub.2 partition wall 10b. Therefore, the oxygen concentration in the first chamber 4 is apt to increase in accordance with increase in separating distance from the pumping electrode 16b, as illustrated by a characteristic curve a. On the other hand, as illustrated by a characteristic curve b, the oxygen concentration in the widthwise direction in the first chamber 4 is apt to have a maximum at a position corresponding to the narrow hole 2 through which the measurement gas G is introduced. Therefore, a part of the measurement gas G introduced into the first chamber 4 through the narrow hole 2, especially a part of the measurement gas G passing through a central portion of the first chamber 4 in the vicinity of the measuring electrode 12b is introduced into the second chamber 8 through the narrow hole 6 without being affected by the pumping electrode 16b. As a result, O.sub.2 is excessively introduced, and hence an error occurs in the measured value of NO due to the influence of excessively introduced O.sub.2.